Four dry starts, four different mistakes
The dry start method (DSM) — growing carpet plants emersed in a humid tank for weeks before flooding — is presented as beginner-friendly because "there's no water to manage." I have run four of them and killed carpet in three, each time in a new and educational way. A post-mortem, in the hope someone skips a step.
Attempt one: the misting grinder
Misted twice daily with a sprayer until everything beaded. Root rot within three weeks. Emersed growth wants humid air and moist soil, not wet leaves. Attempt four's rule: mist the substrate line only, never the plant tops, and only when the glass stops fogging by itself.
Attempt two: the airtight terrarium
Sealed the tank with cling film "for humidity." Mold arrived on day nine — white, fuzzy, fast. I lost the front third of a Monte Carlo carpet to it. No seal since; a loose lid or a couple of small gaps gives you 90%+ humidity with enough exchange that the fungi never get the stale still air they want. Since then: zero mold, including in the rainy season.
Attempt three: the early flood
Flooded at day 18 because the carpet looked "filled in." It was not — the spread was a single layer of leaves over an unrooted mat. Half of it floated up within days and the front never recovered. Attempt four ran 45 days before flooding, and the difference at flood time was structural: a mat you can lift with tweezers without tearing.
Attempt four: the boring success
Same Hemianthus callitrichoides, same soil, four changes: mist substrate not leaves, no airtight seal, 45 days minimum, and a cheap hygrometer inside the tank so "humid" stopped being a feeling. Flooded on day 47, zero float-up, full carpet by month three. The light that ran the DSM at 40% now runs the flooded tank on the schedule from the 60P log — the LED notes cover why dimming matters more than fixture choice.
One non-obvious note
DSM plants convert to submersed form after flooding and often shed their first emersed leaf set — same as farm pots. The "it melted after I flooded!" panic two weeks in is almost always the scheduled transition, not a failure.
The method earns its keep on carpet species where submersed growth starts painfully slow. For everything else I now just plant and wait.